Witching Stone Postmortem

Hello, this is the postmortem for my game Witching Stone, which is currently available on Steam and itch.io. The elevator pitch is: match puzzle deckbuilder with anime art.

Some would take issue with this being a deckbuilder as it doesn’t use cards, and there are no retention or shuffle mechanics. But I don’t really want to get into that here, because this document is already very boring.

Stakes

I wanted to spruce this up a bit with some exciting stakes. My previous two games failed to make more than minimum wage, and so if this game fails will it be over for me? But I kept thinking of Indie Game: The Movie, where one of the developers tries to drum up some drama about how if this game fails their career will be over, but undercut this somewhat by having it said over a shot of him nursing a glass in a lavish hotel bar with five chandeliers. And while I am perhaps short five chandeliers, I’m not going to be destitute if Witching Stone fails.

It is true that being able to make fully independent games is something of a privileged position. Self-funding a project obviously requires… funds. And most people do not have the financial security to spend over a year of their life working on something while keeping the lights on and paying others what they deserve. Even if you are going a more traditional route with a publisher or external funding, you’re still going to need to spend months on a pitch and demo and that might not go anywhere.

But, while I do come from a working class family, I was extremely lucky that some stars aligned three years ago with company buyouts, and something I’m not allowed to talk about for legal reasons, that meant I could take some swings at making my own indie games.

What Went Well

This is less a list of just everything that went well, but things I can learn from and try to replicate in the future...

Stuck To A Plan

Before Witching Stone I worked on a different game, which will be in the What Went Bad section, where I dived into production with a prototype toy and just wasted time. With Witching Stone I wanted to make a plan and be sure this is something I could finish.

Here what the game looked like exactly a year before release. And it’s not that different, design wise, to what it was a year later. It even has the same two starting spells!

There were prototypes and experiments! The core mechanic wasn’t revealed to me in a dream and I did have to play around a lot first. Originally it was more of a “duel” where you and a single opponent took turns to draw on the match grid. I also did tests such as a pixel art mock-up to see if this is something I could take on, as I had not done much pixel art before.

Music and Art Outsourcing

Sam Miller made most of the illustration art for the game, including the promotional "key art", and I would say the vast majority of people clicked on its page not because of anything the game itself was doing, but because of the 90s anime aesthetic reminding them of Slayers or Saber Marionette. I guess I did some art direction, in that I had to describe how characters look in the game, but I’m not going to overstate my influence! He has a whole thread on this here.

I've been helping my friend Alex in my spare time with art for his new game Witching Stone! I took inspiration from artists I admire such as Tsukasa Kotobuki, Rui Araizumi and Kunihiko Tanaka in creating some nostalgic feeling anime visuals. 🧵 pic.twitter.com/1jkQKt66AC

— Sam Miller (@Underbirddd) June 3, 2024

For music I approached Noelle Aman asking if she would be interested. I was first introduced to her work via her Youtube channel on visual novels and eroge (my journey started with The Wild Barcode Battling Trans Rights Mecha Shonen & Its Doujin Sequels!) and found out she’s also a very talented composer. I’ll admit I don’t know much about making music and the Black Books line “I must be musical, I have hundreds of CDs” did come to mind a few times when failing to describe what the game wanted, but that's why you ask people to help!

Despite asking for 9 different battle tracks I think they all sound distinct and have a nice ramp from the five regular tracks to boss, to bigger boss, to a bigger boss again, to an even bigger boss, to okay this guy is definitely the final boss. The shop and environment are also great. It’s all great. I’m particularly fond of the Abyss track because graphically that area is very minimalist and it does most of the heavy lifting when setting the vibe.

Also, you can buy the soundtrack. Please go and buy the soundtrack! It all goes to her. Well, and Bandcamp if it’s not The Special Money Day.

In addition to bespoke help, I also bought a lot of assets on itch.io and Booth. Including a lot of VFX and my first set of enemies, and I don’t regret a penny! I also used free assets for fonts and buttons, etc. Thanks to everyone who lets me do this!

Creating Pixel Art

This was one of the aspects of the game I had to be realistic about and come up with a plan. The original mock up by Sam looked great, but there was no way we could collaborate on this scale. I do not have the money to hire him full time and... he didn’t want to be hired full time anyway! He’s got a job already, thank you very much!

So I decided it’s time to learn a new skill and attempt some pixel art, while he helped out with portraits, some general art tips, paintovers, etc. mostly just as a friend who happens to be an extremely talented artist.

I am, broadly speaking, happy with how it turned out. I don’t think I am amazing at pixel art now, and it is perhaps a bit too derivative of my inspirations (especially Momodora), but I did manage to make a full game with it and people have been nice about it!

All my characters do have a “guys, I need to pee so bad” bounce to them as I used the same technique to add movement to every idle, but it’s way better than I thought it was going to be!

Game Balance

The idea that balancing went well is probably of some dispute. Not even six hours after the game came out I received a negative review from a balance expert who wondered if I had even played my own game.

Balance is something that we talk about a lot as both game creators and players, typically in an argumentative style where you say something subjective so confidently it makes it objective. What “balanced” means varies massively between game types and player preference. And I think if you want a meaningful discussion you first need to define, and agree upon, an idealised version of the game.

As an example: A competitive multiplayer game might set its design goal to have every character sit on an even win-rate. That game could then bring up a match-up chart to decide who needs to be hit with hammers. I’m not sure this actually makes the best experience, it encourages a more conservative design where everything is considered as a zero sum trade off from a template, and ignores that people love underdogs and heels... but that’s a discussion for another time.

I think this idealised version is particularly hard to nail down in deckbuilders where going exponential and exploding a boss is considered, with multiple vague caveats, very desirable.

So, all I can really do is tell you what I find fun in a roguelite and tried to encode in my game. Which is the satisfaction of putting together a working build from disparate scraps and mechanical knowledge. I wanted this to be a swingy experience, where sometimes the player needs to choose between two subpar things and it’s a sweat, and other times everything lines up perfectly and they two-shot the boss. And I believe the latter is meaningless without the former. What I consider undesirable is for a run to be utterly doomed from the start due to generation on a difficulty of +2 or lower, and some starting options (ex: Dancer) probably do fall short here.

In fact, a day after writing this section, the Daily Challenge did offer up a starting option that was probably impossible and I hotfixed it. This flaw was more in how Spells were picked for the Daily Challenge rather than the data itself, but still a huge mistake and something that, admittedly, might happen again. I’m not going to cut and paste this into the “What Went Bad” as a result, but it is an example of The One Thing We Didn’t Want To Happen.

Builds also don't exist in a vacuum and the combat itself needed a lot of variables in what enemies do. Something I kept thinking about here is Slice & Dice, where there are 5 different possible bosses and Hexia and The Hand require specific counters. A player can get there without an answer and just die. But is that unfair? There was 20 rounds to prepare for this. So, I don't think it's a problem that Mother punishes having no backline access or ways to deal with plate.

Another potential problem is the match board just being too random. What if nothing usable comes up round after round and you lose? The game does not do any shenanigans based on board state, but it does use a "stone bag" similar to Tetris pieces to determine what comes in, making sure the average stone distribution is fair even in the short-term. I am sure people have died because their board was garbage, but I think board management is a skill and this is rare outside of the Morgen fight (who does have some politeness code - you'll notice she never walls off the stage vertically).

An example of something that was not a concern is making sure every option has an even “competitive win rate.” And I do not mind that Arawn, a character unlocked by beating a hidden superboss only hinted at during the also-hidden true final boss, is harder to play than the starting character. Judging it on those terms is still fine if that’s your preference, you don’t have to like my design aspirations just because I've thought about them for five minutes, but there is method to the madness.

I also wanted to celebrate diversity and have players feel like they can experiment with different options without the game judging them. So each spell, badge and modifier gets a little ribbon if used for completion, but I do not show win rates or win streaks. Players shouldn't feel like they have to take Famine or Engine each run because they are strong, as losing doesn't matter so who actually cares.

With that all ass-covering out of the way, I think I'm broadly happy with the balance in Witching Stone and I don’t think it’s a cope to say the nature of the game means imperfections make the game more interesting. I have worked on competitive multiplayer titles where I would want sheets of analytics to look at, but I don't feel like it's needed here. Which is good because I don’t have a solution for that.

Code and Bugs

I usually refer to myself as a designer rather than a programmer but I guess I have coded multiple games now! I think some genres need a programmer/designer hybrid who understands the mechanics down in the metal. One example is fighting games that operate on the level of bounding boxes and animation frames. And I think deckbuilders are in a similar space, although for different reasons, as often a turn in a deckbuilder is multiple different scaling systems and combat mechanics colliding at exactly one point in time. I don’t think I am a particularly amazing programmer, but this particular intersection is my jam.

Witching Stone is not perfect. There are some problems with at-a-glance knowing how something will work. The order of operations for defence is consistent but black box, rounding is unclear, etc. I would say Wildfrost and Slice & Dice both do this much better. But it is not just stuff going off willy-nilly in a big combat soup.

Because of my lack of professional programming experience I expected the game to be a bit of a hot mess upon release. But outside of one crash bug there was nothing serious. Most of the bugs found would have come up in a day's testing by any QA department but unfortunately I didn’t have the luxury of one of those.

So, yeah, there are still a lot of minor bugs and I’m hoping to get them ironed out, and again I am not claiming the game is perfect and free of any issues, but it seems surprisingly stable and I’m happy given my lack of experience and its complexity.

Game Demo

I feel like the biggest success for Witching Stone was the public demo. I should stress that this demo wasn’t a prototype or a marketing "vertical slice." It was mechanically complete, reasonably polished, had tutorials, had a full act of replayable content with three bosses, and two playable characters. This was functional in-browser on itch.io meaning it had very few barriers to play and it got a lot of traffic. This was also very motivating for me as people liked the demo a lot. I think I would absolutely go for this route again in the future if viable.

As I launched the page on itch.io I also launched the Steam page so I could forward people as a call to action. I also popped the demo onto Steam a little bit later for Next Fest where I think it was a lot more popular? Japan is my top sales country by far and itch.io is a very English-focused website from my experience.

I also took the demo to Tokyo Game Dungeon for feedback! Tokyo Game Dungeon and Yokohama Game Dungeon are doujin game events here in Japan. It’s first come first served and you get a table, monitor, extension cable and chair, and you can show your game off to anyone who walks by.

I’ve been to these events as an exhibitor and player and I find them really fun and unique. Exhibiting at a show in London means paying through the nose and/or passing some indie approval process to be placed in the shadow of a 12ft Devolver Digital mascot. But Tokyo Game Dungeon has very low barriers, the price is cheap and has no selection process, so if you’re a hobbyist making a game about ear wax removal or a free game about Usada Pekora, you can just show up and people can see your game. And I think that is very cool!

Honestly, I don’t think it really means much in terms of sales. I did shift a couple hundred flyers across two events, but I have no idea how many people went to buy the game and I suspect it’s not many! But it was really useful and motivating to just see normal people playing and enjoying the game, and publishers who talked to me did seem genuinely interested.

Youtube and Streamers

I have seen a lot of articles along the lines of “a youtuber played my game and got 3 trillion views but I only sold 2 copies.” And I have to say this is the opposite of my experience. For Witching Stone, and indeed all my games prior, my wishlists and sales graphs all have little mountains on them that I can name after individual creators like Mt. Wanderbots, Mt. Olexa or ハヤトの野望山.

Sam and I are big fans of The Infinite Review and he submitted it to the people's E3 they do each year. It appearing in there was super fun and "This is E3" was one of the first reviews I received! This stuff helps! I think they edited out the boobed-plant from the trailer but that's okay!

Perhaps because my problems are more discovery related (more on this later, though!), and not that they are in an overcrowded genre or easy to spoil in a video? Maybe more likely is that coverage has usually been from channels whose audience are gamers looking for games to play, rather than general "entertainment" channels.

What went bad

I would like these to also be things I could learn from in future projects...

Roguelink

This is kind of the inverse of the "Stuck To A Plan" section. Before Witching Stone I was working on another game, working title Roguelink, which did not work out. I had a fun prototype, and I still think the core is worth revisiting in the future, but didn’t really know how I could turn it into a full game.

This means I did not know how I would create (or source) all the assets I needed, and did not know how gameplay could be expanded to a full game. I also had trouble making player builds fun. They rarely affected gameplay, and had limited vectors in terms of inputs/outputs, and that is basically a roguelite crime.

Aware I was using up time, I decided to just start production anyway hoping I would figure this out and I didn’t. Half a year later I had to admit this wasn’t working, and it took me another 3 months of experimentation to stop. This meant that I had effectively burnt through 9 months with nothing much to show for it.

I'm worried this sounds a bit soulless. It’s great to jump in, jam, and splash about. But at some point you need to sit back, grasp the full picture, and make a realistic plan with a reasonably solid vision. This doesn’t mean listing out the stats of every enemy ahead of time! It means knowing your base systems, how they plug together, and roughly the content scope for something you could ship and feel good about.

I have worked on projects where, two or three years in, you could ask the lead what we’re making and they couldn’t tell you outside of an elevator pitch. Sometimes not outside of an elevator word like “esports.” Hobbling along directionless hoping it can catch some key feature or current trend and save itself. This is a trap for complacent, successful studios and just death for indies.

Secret Characters

Witching Stone has six playable characters but on launch three of them were completely hidden. Unlocking these characters meant doing something strange like customising a spell to have a null cast sequence, which causes it to fake-glitch out, and then casting it multiple times until it summons a boss.

The assumption was knowledge about these characters would spread, someone would make a Steam Guide, people would mine achievements or see the gaps in their spellbook. But weeks after release Steam achievements suggested less than 5% of people have unlocked one of these characters. People even left positive reviews saying they’d like a 4th character at some point!

In my first big patch I made all the secret characters appear on the character select screen with cryptic clues on their unlock criteria just so people knew they existed. And my ultimate takeaway is the bonus characters should not have been this secret.

I also think a more business savvy person would have launched with three base characters and then, every month after release, do a big announcement about the Pwca Patch or whatever and turned another one on. It’s a big update, reengages players, makes the game look more alive, etc. This is definitely something I see with Vampire Survivor-Likes and... I guess it kind of makes sense.

Post-Release Support

I feel like there is an expectation these days that after a game launches there will be a roadmap of future features. Even an indie game not in early access, made by a single developer, that has flopped big time. Steam itself encourages this with additional Visibility Rounds, the only way to boost your game post-launch, requiring a significant update (undefined) as a prerequisite.

After I launched SOKOBOT and Touhou Library Survivors I did have ideas for content I wanted to produce under the proviso that those games had an active base. Neither of them really did, so I did one big content update for both, adding the Ribena blocks to SOKOBOT and Flandre Scarlet to Touhou Library Survivors, and I basically moved on.

Touhou Library Survivors in particular had a lot of space left to explore because the expansion opportunities are obvious: Add more Touhou characters! I could add a new stage based on PCB (next in sequence) or MoF (my favourite). I really wanted to add Kawashiro Nitori as playable as the game was originally about her. But it was hard to justify! These double stage or character content.

With Witching Stone I have a different problem. While the player base is declining, I probably can still justify adding a bit more content to the game past what I’ve done so far (Daily Challenges). But I actually don’t have a big well of ideas to return to! I launched with three more characters than expected and tried to leave nothing on the table mechanically. So a lot of the things I’m trying seem gimmicky and forced, and one of the games biggest compliments is that it’s quite lean in scope.

$17.99

The game was $18 dollars base and released at $15 dollars due to the discount. The actual base $18 figure came from me reading a post by Simon Carless on indie game pricing from a few years ago. Where he argues a game of sufficient size should cost $20 or more. I was convinced! And then came back a bit later and knocked it down two tiers because, you know, calm down. There was a lot of anxiety around this pricing. But I had justified this to myself. And I will now give you two such justifications:

The first is that the price is pretty sandwiched. There are more expensive games like Slay the Spire, Monster Train, Inkbound and Demon’s Mirror, but also cheaper games like Balatro, Dicey Dungeons, Dicefolk and Slice & Dice. While I admit many people will enjoy these cheaper games more than Witching Stone, I don’t think most potential customers are aware of an objective, linear relationship between game price and their personal enjoyment, and I reasoned Witching Stone would be more niche due to its matching mechanic, so I was okay to aim a little north.

Secondly, this game has been mostly promoted via a playable demo that has been constantly updated for the last half a year. All coverage up to release, with one exception, had been on content available for everyone to play, and this could be considered a premium upgrade for a demo people are already sinking 25 hours into. So I am not asking players to take a gamble if they don’t want to.

Okay, that was my justification. What do I actually think now? It probably should have been cheaper. While I haven’t had any negative reviews complaining about the lack of “bang per buck” I do think it’s preventing interested people from taking the plunge. Some potential clues for this are in my wishlist conversion rates:

• SOKOBOT: 20.7%
• Touhou Library Survivors: 31.8%
• Witching Stone: 6.3%

I still roughly agree with my logic above. But I think the amount of high quality games released at a low price point has anchored the price for a deckbuilder to be around $12, unless you have some serious prestige or marketing chops. And let’s face it, I don’t have any chops. I also see a lot of people recommending trying the demo, which is cool, but I wonder if it was cheaper they'd just say buy the game.

However, I only saw one person actually complain about the price being too high, and they were a troll on my discussion boards, and I would be making less per sale. So... I also don't think this is that dramatic. I don't think in that alternate timeline I would now be buying a superyacht.

Revenue

I am writing this 2 months since release as I wanted to do at least one discount sale first. It has made just under $30,000. But this is before Steam take a third, and before expenses such as art, music, event gear, equipment, etc. It is also going to sell less and less as time goes on, and has already dipped to 1-2 sales on multiple days. So, I haven't made Balatro here!

I want to be careful with my words here because I know a lot of indie games do not make this much. This puts me in a shockingly high percentile of indie success. And I don't want to sound like this was a purely financial venture as the main draw to "going indie" was making the games I want to make... Howeveeer, ideally this would generate the equivalent of two years of my old game industry salary and I just don't think that's possible. I live in hope it will catch some spark and "pop off" but that's not something I am really in control of. I can't make Northernlion play my game.

Discovery

There's this image that floats around like "indie tips: get 10,000 wishlists" which is kind of annoying because it's like "tips for success: be successful." But this time I did crack 8,000 before release and get in Featured Upcoming. And that's pretty good!

"Build it and they will come" is obviously not true because we invented this evil we call marketing. With Witching Stone I tried a new approach of leading with the online demo, and I felt like I had a lot more social media presence than before, but let's be honest I'm mostly followed by other developers and not potential customers, and I've never had followers in 4 digits. I think a lot of this is me doing better, but not good.

Ultimately my biggest sale region is Japan and while that is cool... it does make me wonder how impactful these things actually were. Because I didn't promote the game in Japanese outside of Tokyo Game Dungeon. I suspect Gwen looking a lot like Lina Inverse is the main reason.

I'll be honest, a lot of the advice on marketing I really don't like. It's "post on reddit under a false pretense, don't link your game but wait for someone to ask, or they'll tear your shilling ass apart" and that makes me feel gross. I am not going to do this but that means I will sell less.

Make TikToks... No. I will not.

Isolation and overwork

I worked on Witching Stone from morning to 2am every single day, including weekends, for a year and a half. My worst industry job (do not work for Climax Studios) forced us to work until 9pm almost every day and one/two weekends a month as the “mini-crunch to avoid the big crunch” (spoilers, it did not, but thankfully my contract expired before that). This was worse than that. The thing is it felt different because I was working on my own thing, that I had a lot of passion for, and not one of the worst action games ever made. But I know this is an insane way to make a game. I know that taking the time to rest will create better results in the same amount of real time.

Part way through this project I moved to Japan because my wife got a new job on the requirement she move back here. As I am working for myself on a computer this seemed like a good opportunity so we went for it. Being in a foreign country where I didn’t know the language or have reliable income definitely gave me a lot of anxiety, on top of my just being autistic and work obsessed. But sometimes I would go outside and literally think to myself “oh yeah, I’m in Japan?” That seems quite bad.

This is the end

I feel like I need some kind of "I liked making this :)" here as a palette cleanser